380 ALASKAN GLACIER STUDIES quoted on a previous page and in our earlier publications, that the ice in the axis of Icy Bay retreated six or seven miles from 1794 to 1908, for this involves the participation of Tiger Glacier in the expansion of 1794. As we revise this text, however, we have the later statement, published in 1911 by Grant and Higgins,1 that in 1794 "it is very probable that the glaciers in Nassau Fiord . . . completely filled that fiord and extended out into, but not across, the main part of Icy Bay." If this means that Tiger Glacier did not participate in the expansion of the Chenega-Princeton ice tongue it is. quite in accord with the distribution of vegetation. One of the most interesting features in relation to forest growth in this region is the shape of the barren zone on the peninsula east of Nassau Fiord, between that indentation and Icy Bay. This shows clearly that while the expanded Chenega and Princeton Glaciers filled Nassau Fiord and extended over the narrow northeastern portion of the peninsula, the higher southeastern side of the peninsula and the low island there, which are thickly forested, were not covered by the glacier. The dashed line (Fig. 56), showing the boundary of the barren zone indicates, therefore, that the expanded Chenega-Princeton Glacier was tidal in Icy Bay in the cove east of this peninsula as well as opposite the mouth of Nassau Fiord; but that the two tidal termini did not coalesce and transform the higher part of this peninsula into a nunatak, nor override it completely. Glacial Modifications in Icy Say. Icy Bay, like the other fiords of Prince William Sound, has been intensely glaciated, giving it very steep walls and large areas of bare, striated rock often having the roches moutonn^e form. Hanging valleys are also present,, the one containing the small ice tongue south of Tiger's Tail Glacier forming an especially good illustration. Icy Bay, even more than the fiords to the north, is remarkable for the absence of glacial deposits above sea level, the chief deposits observed by us being scattered, glacial bowlders on the bare rock surfaces and a thin veneer of ground moraine in places. There is a narrow strip of outwash gravel which separates Princeton Glacier from the fiord, and a very perfect, though minute, terminal moraine ridge, in the second cove east of Nassau Fiord on the northern side of Icy Bay (a, Fig. 56). It is made up of stony till and extends across the valley, at the head of the cove, as a single, narrow-crested ridge, with a height of between 20 and 30 feet. It seems to have been built by a projecting lobe from the Icy Bay Glacier rather than by a retreating glacier in this small valley. GlACTEHS ON THE ISLANDS OF PRINCE WlLLIAM SOUND Knight Island. Knight Island, in western Prince William Sound, east of Icy Bay and Port Nellie Juan, is 26 miles long and 7 or 8 miles wide (PL XCIEE). Its coast is deeply indented by fiords and its surface is very rugged. The peaks are 2000 to 3180 feet high, and since many of them are snow-capped, it is probable that their nev6 fields give rise to a number of minute glaciers. None of these have been shown on a map, and only one of them, a cascading glacier on the southern side of the island near Mummy Bay, was seen by the National Geographic Society's expedition. The large number of cirques show clearly that extensive local glaciers have existed in the past. Their relationships to the expanded ice sheet of Prince William Sound are discussed in a later chapter. Chenega Island. Chenega Island, between Knight Island and Icy Bay, is about 8 > Bull. Amer. Geog. Soc., Vol. "gT.TTT, 1911, p. 414.